Few topics in the Mac universe engender more confusion than market share numbers. The almost infinite ways to measure market share – who’s buying, where they’re buying (Internet vs. brick and mortar), laptops vs. desktops, aggregate metrics from Web sites, raw sales data from all the PC manufacturers – make any accurate assessment almost impossible.

So I was amused yesterday when many Mac Web sites celebrated the latest numbers from California-based Net Applications, which each month measures market share of Web browsers and operating systems. The company’s September data shows Mac OS Web traffic at 6.61 percent (a combination of 3.38 percent for PowerPC-based Macs and 3.23 for Intel-based Macs). Everyone seemed pleased that the Mac OS share had risen almost a half a percentage point from 6.15 percent in August and more than a point from 5.21 percent last October.

But if you look at the data for each month over the past year, the achievement looks less impressive. The Mac OS had a combined share of 6.22 per percent as early as January. It has hovered between 5.97 percent (July) and 6.46 percent (May) all through 2007. Perhaps the best thing that can be said for the September number is that it is the highest so far this year, but by a statistically insignificant amount.

The numbers for Safari, Apple’s Web browser, are similar. Safari’s share edged up to 5.07 percent from 4.68 percent in August. Last October Safari had just a 3.93 percent share of Web surfers. But Safari had 4.85 percent in February and 4.82 percent in May so September’s jump, while the highest of 2007, does not represent a significant gain in market share for Safari.

Earlier today an outfit called W3Counter released comparable statistics showing the Mac OS representing a mere 3.77 percent of Web traffic. That’s unchanged from W3Counter’s August number, and a slight drop from the 3.91 percent the Mac OS recorded in May. In the browser category, Safari had a lowly1.77 percent share.

Why the discrepancy? Apparently the two companies are measuring traffic on different Web sites, though nether gives a lot of details on which sites they measure. Since both are in the business of supplying traffic data to customers, one assumes the statistics come from each company’s respective roster of clients. If true, that’s not a scientifically reliable way to sample market share of operating systems and browsers.

For its part, W3Counter provides evidence that their data is skewed toward the unreliable. The nation category shows tiny Latvia as the country generating the fourth-most Web traffic, ahead of Canada, the Philippines, France and Australia.

Taken individually, no data on Mac market share can offer definitive evidence of the platform’s progress against Windows. Nevertheless, most surveys and reports released over the past year or so have shown the Mac making headway, albeit slowly.

We’re unlikely to see any sudden, dramatic surges in Mac market share over the next year or two – Apple’s disinterest in pursuing the enterprise (business) market precludes any large gains -- but expect to see gradual growth as more and more consumers look to the Mac as a viable alternative to Windows.

What would be really interesting is a survey showing how the Mac’s market share has increased among consumers, omitting all those Dells in the workplace. Any takers?